Date: Sat, 9 Jan 1999 11:43:33 -0500 From: "Gene W. Homicki" <gwh@spiders.com> To: Nate Williams <nate@mt.sri.com>, Luigi Rizzo <luigi@labinfo.iet.unipi.it> Cc: java@FreeBSD.ORG Subject: Re: compiler speeds Message-ID: <19990109114333.M20734@spiders.com> In-Reply-To: <199901061626.JAA13931@mt.sri.com>; from Nate Williams on Wed, Jan 06, 1999 at 09:26:58AM -0700 References: <199901060903.KAA05333@labinfo.iet.unipi.it> <199901061626.JAA13931@mt.sri.com>
next in thread | previous in thread | raw e-mail | index | archive | help
+---
| > class files slightly differ in size. I cannot tell about the quality of
| > the code.
+---
There is a bit of information on the AlphaWorks site
(alphaworks.ibm.com) about some of the optimizations that Jikes does
The information isn't all that complete last time I looked. The source
code to Jikes can of course be perused now. :-)
As for real-world speed... doing a clean build of one of our projects
(about 850 .java files to compile):
with javac (no JIT):
1401.93 real 1261.02 user 32.17 sys
with jikes:
112.32 real 45.05 user 14.61 sys
I love jikes.
+---
| So far I've been using jikes to develop 'production' code, and given
| that there aren't any significant optimizations in either javac or
| jikes, I feel pretty safe with it now.
+---
The only thing to be careful of is if your code needs to be deployed on
a Java 1.2 (the "Java 2 Platform") system. Last time I tried
it, the Jikes compiled code didn't pass the 1.2 verifier.
Gene
--
Gene W. Homicki gwh@spiders.com
Objective Consulting, Inc. http://www.spiders.com/
Internet Application Design voice: +1 914.524.4151
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-java" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990109114333.M20734>